Sleep stage classification algorithm

Abstract

Methods, systems, and devices for a sleep staging algorithm are described. The system may receive physiological data associated with a user from a wearable ring device, where the physiological data is collected via the wearable ring device over a time interval. The system may input the physiological data into a machine learning classifier and use the machine learning classifier to classify the physiological data into at least one of a set of sleep stages for at least a portion of the time interval. The system may then cause an indication of the at least one of the set of sleep stages to be displayed on a graphical user interface (GUI) of the user device based on classifying the physiological data.

Claims

1. A method for automatically detecting sleep stages, A step of receiving physiological data associated with a user from a wearable ring device, wherein the physiological data is collected via the wearable ring device over a time interval, The steps include inputting the aforementioned physiological data into a machine learning classifier, A step of classifying the physiological data into multiple sleep intervals within the time interval using the machine learning classifier, wherein the multiple sleep intervals represent the duration of time the user was asleep. The steps include using the machine learning classifier to classify each of the multiple sleep intervals into the corresponding classified sleep stage of a plurality of classified sleep stages, The steps include displaying one or more sleep intervals from the plurality of sleep intervals, and one or more classified sleep stages corresponding to the one or more sleep intervals, on the graphical user interface of the user device. Methods that include...

2. The aforementioned multiple classified sleep stages include at least one of the following: wakefulness, light sleep, rapid eye movement sleep, or deep sleep. The method according to claim 1.

3. For each of the one or more classified sleep stages corresponding to the one or more sleep intervals, at least one of the following is displayed: the duration of each classified sleep stage, or the percentage of the total sleep time spent by the user in each classified sleep stage. The method according to claim 1.

4. The step further includes performing one or more normalization procedures on the physiological data, and the step of inputting the physiological data into the machine learning classifier includes the step of inputting the normalized physiological data into the machine learning classifier. The method according to claim 1.

5. The step of classifying the physiological data further includes using the machine learning classifier to identify a plurality of features associated with the physiological data, wherein the step of classifying the physiological data is at least in part based on identifying the plurality of features. The method according to claim 1.

6. The aforementioned features include the rate of change of the physiological data, the pattern between two or more parameters of the physiological data, the maximum data value of the physiological data, the minimum data value of the physiological data, the mean data value of the physiological data, the median data value of the physiological data, a comparison of the data value of the physiological data with the user's baseline data value, or any combination thereof. The method according to claim 5.

7. A step of displaying one or more of the aforementioned features on the graphical user interface of the user device. The method according to claim 5, further comprising:

8. A step of identifying the bedtime associated with the user, the wake-up time associated with the user, or both, at least in part on classifying the physiological data, The steps include displaying the bedtime, wake-up time, or both on the graphical user interface of the user device, The method according to claim 1, further comprising:

9. The step of inputting the physiological data into the machine learning classifier is: The step includes transmitting the physiological data to one or more servers for classification via the user device, The method according to claim 1.

10. The method further includes the step of using the user device to generate one or more scores associated with the user, at least in part, based on the physiological data, wherein the one or more scores include a sleep score, a readiness score, or both. The method according to claim 9.

11. The step further includes inputting a circadian rhythm adjustment model into the machine learning classifier, wherein the step of classifying the physiological data is at least partially based on the circadian rhythm adjustment model. The method according to claim 1.

12. A step of receiving additional physiological data associated with the user from the wearable ring device, wherein the additional physiological data is collected via the wearable ring device over a second time interval. The steps include inputting the aforementioned additional physiological data into the machine learning classifier, A step of using the machine learning classifier to classify the additional physiological data for at least a portion of the second time interval into at least one of the plurality of classified sleep stages, wherein classifying the additional physiological data is at least in part based on inputting the physiological data and the additional physiological data. The steps include, at least in part, displaying instructions for at least one of the multiple classified sleep stages within the second time interval on the graphical user interface of the user device, based on classifying the additional physiological data, The method according to claim 1, further comprising:

13. A step of displaying at least a subset of the physiological data on the graphical user interface of the user device, The method according to claim 1, further comprising:

14. The physiological data includes body temperature data, accelerometer data, heart rate data, heart rate variability data, blood oxygen concentration data, or any combination thereof. The method according to claim 1.

15. The wearable ring device collects physiological data from the user based on arterial blood flow in the user's finger. The method according to claim 1.

16. The wearable ring device collects the physiological data from the user using one or more red light-emitting diodes and one or more green light-emitting diodes. The method according to claim 1.

17. A device for automatically detecting sleep stages, Processor and The memory coupled to the aforementioned processor, Instructions stored in the memory, which the device has Receiving physiological data associated with a user from a wearable ring device, wherein the physiological data is collected via the wearable ring device over a period of time. Inputting the aforementioned physiological data into a machine learning classifier, The process involves classifying the physiological data into multiple sleep intervals within the time interval using the machine learning classifier, wherein each sleep interval represents a period of time during which the user was asleep. Using the aforementioned machine learning classifier, each of the multiple sleep intervals is classified into the corresponding classified sleep stage of the multiple classified sleep stages. Displaying one or more sleep intervals from the aforementioned plurality of sleep intervals, and one or more classified sleep stages corresponding to the one or more sleep intervals, on the graphical user interface of the user device, Instructions that can be executed by the processor in order to perform the following: A device equipped with the following features.

18. The plurality of classified sleep stages include at least one of the wakefulness sleep stage, the light sleep stage, the rapid eye movement sleep stage, or the deep sleep stage. The apparatus according to claim 17.

19. For each of the one or more classified sleep stages corresponding to the one or more sleep intervals, at least one of the following is displayed: the duration of each classified sleep stage, or the percentage of the total sleep time spent by the user in each classified sleep stage. The apparatus according to claim 17.

20. The aforementioned instruction is given to the device, Perform one or more normalization procedures on the aforementioned physiological data. Further execution by the processor to perform the following is possible, and inputting the physiological data into the machine learning classifier includes inputting normalized physiological data into the machine learning classifier. The apparatus according to claim 17.

Images